1. Field of the Invention
The present invention relates to image pickup apparatuses such as digital still cameras, and particularly relates to a technique of storing and managing data representing an image captured using an image pickup apparatus.
2. Description of the Related Art
As a format of images captured using digital still cameras, a DCF (Design Rule for Camera File System) is established by JEITA (Japan Electronics and Information Technology Industries Association). The DCF defines a directory configuration and a file format used when data representing an image captured using a digital still camera is recorded in a removable memory such as a CompactFlash (registered trademark).
In the DCF, the file configuration is defined by an Exif (Exchange Image file Format). Furthermore, the image data is compressed by a JPEG (Joint Photographic Expert Group) method, and thumbnail images, each of which has a pixel matrix of 160×120 pixels without compression, are obtained. Furthermore, camera information may be recorded when the image is captured along with the image data by an Exif tag.
Here, the camera information includes a date, a name of a manufacturer of a camera, a model name, a software version, a shutter speed, an aperture, an auto-exposure program, and a description about a photograph. In such camera information, information other than the description about a photograph can be automatically obtained when the digital still camera is manufactured or when an image is captured.
In the camera information, detailed unique information relating to an object or a shooting location is generally described as the description about a photograph. Examples of the description about a photograph include a description about what is captured and a description about where the image is captured, such as a description “2007 school year XXX elementary school entrance ceremony” and a description “XXX aquarium dolphin show”.
To automatically obtain such detailed unique information on a photograph is considerably effective in terms of operability. In order to automatically obtain such detailed unique information on a photograph, for example, automatic GPS (Global Positioning System) information may be utilized or an image recognition technique may be utilized.
However, although general geographic information such as information on Shinjuku in Tokyo can be obtained in terms of a location and match or mismatch of faces can be recognized by the GPS information or the image recognition technique, detailed unique information required for managing photographs (image data) such as names of sightseeing spots, amusement parks, museums, and theme parks cannot be obtained. Therefore, in the related art, detailed unique information on data representing a captured image, that is, description data cannot be managed by automatically associating the description data with the image data.
To address this problem, Japanese Patent Laid-Open No. 2003-198909 discloses a technique of storing in advance a file which stores description data describing an object along with an index in a memory card and inputting the index before an image is captured or after an image is captured. Specifically, obtained image data is stored in the memory card in a file format, and a file regarding the description data is associated with a file regarding the image data using the index to be input.
Furthermore, Japanese Patent Laid-Open No. 2003-242440 discloses a technique of extracting text regions from recognized objects which are included in a plurality of images having different brightness levels included in a moving image, and recognizing characters included in the text regions by means of matching processing performed on the text regions using dictionary images. Specifically, among results of pattern recognition performed on the plurality of images having different brightness levels, a result which represents a high matching rate in the matching processing is output as a final result of the pattern recognition. When this technique is used, text which is difficult to be recognized in an environment which is too dark due to a shadow of an object near the image in the daylight can be appropriately subjected to the pattern recognition using the plurality of images having different brightness levels. Furthermore, text which is difficult to be recognized in an environment which is too bright due to illuminance near the image at night can be appropriately subjected to the pattern recognition using the plurality of images having different brightness levels. Accordingly, accuracy of the pattern recognition is improved.
However, in the technique disclosed in Japanese Patent Laid-Open No. 2003-198909, although the description data which describes the object should be stored in the memory card in advance, the description data cannot be input unless information on a shooting location, for example, is determined. Therefore, when the information cannot be specified, the information is required to be input from the camera or a PC, for example. In this processing, an operation of inputting characters is considerably troublesome when a device which has a limited inputting method such as a camera is used.
Moreover, the technique disclosed in Japanese Patent Laid-Open No. 2003-242440 is used for optimizing text recognition in accordance with change of brightness depending on an environment of shooting of a moving image and is not used for managing information on a shooting location, for example, by automatically associating the information with image data.